Scholarlink

<p>Current keypoint-based approaches used in Unmanned Aerial Vehicle (UAV) photogrammetric surveys cannot control the error propagation effectively due to the sub-optimal number of Ground Control Points (GCPs) placed within the surveyed area. Spurred by this issue, this paper introduces a marker system that provides many precise projections of artificial markers as point feature constraints, thereby improving the accuracy of UAV-based photogrammetry. Specifically, we introduce a marker system with colored circles and white backgrounds and the corresponding detection algorithms to quickly suppress background interferences using different color spaces for accurate marker localization. The first experiment demonstrates that the proposed marker system has a better detection performance than current marker systems used in aerial surveying projects. Additionally, a second experiment reveals that, in cases with only 4 GCPs, the proposed marker system significantly improves the accuracy of the UAV-photogrammetric survey. Compared to Agisoft Metashape, adding additional point feature constraints to the markers increases the accuracy of Check Points (CPs) in the X, Y, and Z directions by 37.5%, 23.0%, and 78.2%, respectively.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Improvement of UAV-Photogrammetric Survey Accuracy Using Point Feature Constraints of Markers

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月22日

查看全文
  • Lei Xing,
  • Yufu Li,
  • Chonglun Yue

摘要

Current keypoint-based approaches used in Unmanned Aerial Vehicle (UAV) photogrammetric surveys cannot control the error propagation effectively due to the sub-optimal number of Ground Control Points (GCPs) placed within the surveyed area. Spurred by this issue, this paper introduces a marker system that provides many precise projections of artificial markers as point feature constraints, thereby improving the accuracy of UAV-based photogrammetry. Specifically, we introduce a marker system with colored circles and white backgrounds and the corresponding detection algorithms to quickly suppress background interferences using different color spaces for accurate marker localization. The first experiment demonstrates that the proposed marker system has a better detection performance than current marker systems used in aerial surveying projects. Additionally, a second experiment reveals that, in cases with only 4 GCPs, the proposed marker system significantly improves the accuracy of the UAV-photogrammetric survey. Compared to Agisoft Metashape, adding additional point feature constraints to the markers increases the accuracy of Check Points (CPs) in the X, Y, and Z directions by 37.5%, 23.0%, and 78.2%, respectively.